Page 1 of 1

steady state problem of a non stasionary system

PostPosted: Tue Jan 29, 2013 2:26 am
by wanghao19820707
Dear all:
my equation system is writen in non stasionary form.I carefully follow the example: fs2000ns and its steady state file which uses Y_obs and P_obs with trend and as non stasionary observable variable and Y_obs and P_obs are declared as unit_root_vars P_obs Y_obs also with trend in fs2000ns file and Y_obs=1 and P_obs=1 in its steady state file.

I have four observable variable Y_obs(output with trend) and I_obs(investment with trend) c(consumption divided by output,so no trend) q(house price devided by output so no trend). I set Y_obs=1 and I_obs=1 in my steady state file.
and my data examination shows that oly c contains a unit root.

Now I have two problem:
1.how should I declare in the varobs block ? should I declare Y_obs and I_obs and c all as unit_root_vars ?
2.run rbc_5 the error message is that The steadystate file did not compute the steady state.
the error shows that Residuals of the static equations:
Equation number 9 : -0.0085362
Equation number 10 : -0.0085362

these two equations are:
Y_obs/Y_obs(-1)=da*y/y(-1);
I_obs/I_obs(-1)=da*y/y(-1);

since I set da=exp(x_a+e_a+MU_a) and MU_a=0.0085 and x_a=0 and e_a=0 at steady state I really don't know what to do.

can anyone give some suggestion?

many thanks!

Re: steady state problem of a non stasionary system

PostPosted: Wed Feb 13, 2013 10:57 am
by SébastienVillemot
  • The "unit_root_vars" command is obsolete. Just remove it
  • You should add the option "nocheck" to the "steady" command. Dynare will no longer complain about the steady state equations not being satisfied. Beware, this means that you have to ensure that your steady state is valid (for those variables which have one)